The efficient representation of visual information is important for many image processing tasks, including compression, noise reduction, feature extraction, and inverse problems. Efficiency of a representation refers to the ability to capture significant information about an image using minimal storage space.
For image compression or content-based image retrieval, the use of an efficient representation implies the compactness of the compressed file or the index entry for each image in the database. For practical applications, such an efficient representation can be obtained via an image transform including a digital image filter.